| Genetical analysis of grain yield and agronomic characters
in haxaploid triticale MADUPURI R.NAIDU and K.S. GILL Department of Plant Breeding, Haryana Agricult ural University, Hissar, India. and Punjab Agril. University, Ludhiana, India. The hexaploid triticales, compared with octoploid types, improved cytogenetic stability, fertility and yield potential, and renewed the interest in breeding of this species. The crossing of hexaploid triticales with octoploid triticales or hexaploid wheats presented many potentially useful plant types (QUALSET et al. 1973), while the rapid increase in yielding ability of hexaploid triticales was affected with the development of Armadillo strains in CIMMYT, Mexico (ZILLINSKY 1974). Under certain ecological conditions, the improved hexaploid triticales outyielded the best wheat and barley check cultivars (KROLOW 1977). The present paper, however, reports the results from early generations of crosses between hexaploid triticale strains. The main aim was to evaluate the prepotency of the parents, and to predict the genetic productivity of crosses for further isolation of superior homozygous genotypes in advanced generations based on information of gene action involved in expression of several agronomic traits. Material and Methods The generations P1, P2, F1, F2, B1 (P1 x F1) and B2 (P2 x F1) were evaluated from each of two hexaploid triticale crosses, namely UC 8825 x TL 116 (cross 1) and Beagle x 6T-70 (cross 2). The parental material was highly selected for meiotic stability, adaptability and fertility but genetically diversed for many other attributes contributing to productivity. Each cross was considered as a unit in the planting scheme. The planting arrangement was a completely randomized block design with three replications of single row plots. The rows were 150 cm long and 30 cm apart with 10 cm inter plant spacing. In each replication, 20 plants of parents and F1, 100 plants of F2 and 60 plants of backcrosses were sampled for recording data on heading data, plant height, tillers per plant, spikelets and kernels per ear, 100-kernel weight and single plant yield. The data were analysed by analysis of variance. The heterosis was measured by the magnitude of the difference in mean performance between F1 and better parent (P1 or P2) and expressed as per cent. In genetic model proposed by HAYMAN (1958) the generation means are expressed as the linear function of various gene effects. In the present study, the variation between generation means was partitioned into variation due to additive and dominance effects and that due to deviations from the additivedominance model by fitting unweighted least square regression analysis to generation means. This permitted the test of the adequacy of additive-dominance model and evaluation of importance of additive and dominance effects in their contribution to genetic variation. Results and Discussion Analysis of variance indicated significant differences existed between generations for all characters measured in both crosses. The parent UC8825 was higher performing than TL116 for all traits except kernel weight. Beagle was later in heading but possessed more tillers/ plant, spikelets and kernels/ear and heavier kernels than 6T-70. The F1 hybrids deviated significantly from the desirable parent in both crosses for heading date, number of spikelets and kernels, and kernel weight, and in cross-1 for plant height and grain yield. This heterotic expression of F1 hybrids tended to indicate a certain degree of dominance for those traits. However, the F1 mean value approximately equal to mid-parent value in either of cross for tiller number, indicating the minor importance of non-additive gene action for its genetic expression. In general, the degree of heterotic expression was higher in cross 1 than in cross 2. |
| --> Next |